Streamlining Real Estate Investment Modeling with ChatGPT

This post is a bit of a deviation from my usual thoughts and musings, but it demonstrates a few interesting points I wanted to share.

Lately, I’ve been modeling multiple investment scenarios for various projects. I often need to adjust share classes, GP/LP splits, and introduce waterfalls to see how they impact the overall model.

ChatGPT-4o has been incredibly helpful in this process. Although it sometimes loses context and introduces bugs, it has significantly sped up my ability to understand and adjust models on the fly without making extensive edits in Excel or Google Sheets.

In my experiments, I was able to get GPT to analyze a dataset consisting of 15 years of P&Ls. I provided it with some adjustable variables and then had it export the results (with formulas) to Excel.

Impressed with the results, I decided to take it a step further. On a whim, I asked GPT to generate some JavaScript and a lightweight front end where I could manipulate the variables directly. After applying some Tailwind styles and cleaning up the output, I had a very useful prototype for quickly modeling and adjusting investor returns.

As an added bonus, I added a dynamic chart to show cumulative returns over the investment period.

All in, it took about 1-2 hours to create and iterate on the MVP and another hour or so to integrate it directly into this webpage.

Is this code battle-tested and production-ready?

Probably not.

But it works today and helps me do my job better.

Just a few years ago, creating and debugging something of this scale would have taken more than a week.

As GPTs continue to improve, they are streamlining processes and making it easier for us to tackle complex tasks efficiently, turning what once were formidable challenges into simple routines.

Play around with the calculator below or scroll to the bottom of the page to see the source code.

Investor Waterfall Calculator

Source Code